[openstack-dev] [tripleo] Mistral workflow cannot establish connection

Steven Hardy shardy at redhat.com
Mon Jul 16 12:05:09 UTC 2018

On Sun, Jul 15, 2018 at 7:50 PM, Samuel Monderer
<smonderer at vasonanetworks.com> wrote:
> Hi Remo,
> Attached are templates I used for the deployment. They are based on a deployment we did with OSP11.
> I made the changes for it to work with OSP13.
> I do think it's the roles_data.yaml file that is causing the error because if remove the " -r $TEMPLATES_DIR/roles_data.yaml" from the deployment script the deployment passes the point it was failing before but fails much later because of the missing definition of the role.

I can't see a problem with the roles_data.yaml you provided, it seems
to render ok using tripleo-heat-templates/tools/process-templates.py -
are you sure the error isn't related to uploading the roles_data file
to the swift container?

I'd check basic CLI access to swift as a sanity check, e.g something like:

openstack container list

and writing the roles data e.g:

openstack object create overcloud roles_data.yaml

If that works OK then it may be an haproxy timeout - you are
specifying quite a lot of roles, so I wonder if something is timing
out during the plan creation phase - we had some similar issues in CI
ref https://bugs.launchpad.net/tripleo-quickstart/+bug/1638908 where
increasing the haproxy timeouts helped.


More information about the OpenStack-dev mailing list